hybrid_jp.change_points module#

Chagne point detection using ruptures library.

hybrid_jp.change_points.binseg(arr: ndarray, nseg=3) list[int][source]#

Use Binary Segmentation algorithm to detect change points.

Parameters:
  • arr (np.ndarray) – 2d array of [x,features]

  • nseg (int, optional) – number of change points to return. Defaults to 3.

Returns:

index of change points

Return type:

list[int]

hybrid_jp.change_points.find_shock_index_from_gradnd(gradnd: ndarray) int[source]#

Find the index of the shock from gradient of the number density.

  • If gradnd is 2d then find the peak of the mean.

  • Assume the shock is located at the peak of grad nd

Parameters:

gradnd (np.ndarray) – gradient of the number density.

Returns:

index of the shock.

Return type:

int